Characterization of binding, functional activity, and contractile responses of the selective 5-HT1F receptor agonist lasmiditan

BJP

R E S E A R C H P A P E R

Characterization of binding, functional activity, and contractile

responses of the selective 5

‐HT

_1F

receptor agonist lasmiditan

Eloísa Rubio

‐Beltrán

|

_{Alejandro Labastida}

‐Ramírez

|

_{Kristian A. Haanes}

|

Antoon van den Bogaerdt

|

_{Ad J.J.C. Bogers}

|

_{Eric Zanelli}

|

_{Laurent Meeus}

|

A.H. Jan Danser

|

_{Michael R. Gralinski}

|

_{Peter B. Senese}

|

_{Kirk W. Johnson}

|

Joseph Kovalchin

|

_{Carlos M. Villalón}

|

_{Antoinette MaassenVanDenBrink}

1

Division of Pharmacology, Department of Internal Medicine, Erasmus University Medical Centre, Rotterdam, The Netherlands

2

Department of Cardiothoracic Surgery, Erasmus University Medical Centre, Rotterdam, The Netherlands

3

Research and Development, Déclion Pharmaceuticals, Inc., Marblehead, Massachusetts

4

Euroscreen Fast Services Unit, Epics Therapeutics SA, Gosselies, Belgium

5

CorDynamics, Inc., Chicago, Illinois

6

Lilly Corporate Center, Eli Lilly and Company, Indianapolis, Indiana

7

Research and Development, CoLucid Pharmaceuticals, Inc., Cambridge, Massachusetts

8

Pharmacobiology, Cinvestav‐Coapa, Mexico City, Mexico

Correspondence

Dr Antoinette MaassenVanDenBrink, Division of Pharmacology, Department of Internal Medicine, Erasmus University Medical Centre, PO Box 2040, 3000 CA Rotterdam, The Netherlands.

Email: a.vanharen‐

maassenvandenbrink@erasmusmc.nl

Present Address

Antoon van den Bogaerdt, ETB‐BISLIFE, Heart Valve Bank, Zeestraat 29, 1941 AJ Beverwijk, The Netherlands.

Joseph Kovalchin, Amathus Therapeutics, 700 Main Street, Cambridge, MA 02139.

Background and Purpose:

Triptans are 5

‐HT

receptor agonists (that also

dis-play 5

‐HT

receptor affinity) with antimigraine action, contraindicated in patients

with coronary artery disease due to their vasoconstrictor properties. Conversely,

lasmiditan was developed as an antimigraine 5

‐HT

receptor agonist. To assess the

selectivity and cardiovascular effects of lasmiditan, we investigated the binding,

functional activity, and in vitro/in vivo vascular effects of lasmiditan and compared it

to sumatriptan.

Experimental Approach:

Binding and second messenger activity assays of

lasmiditan and other serotoninergic agonists were performed for human 5

‐HT

, 5

‐

HT

, 5

‐HT

, 5

‐ht

, 5

‐HT

, 5

‐HT

, 5

‐HT

, and 5

‐HT

receptors, and the results

were correlated with their potency to constrict isolated human coronary arteries

(HCAs). Furthermore, concentration

_{–response curves to lasmiditan and sumatriptan}

were performed in proximal and distal HCA, internal mammary, and middle meningeal

arteries. Finally, anaesthetized female beagle dogs received i.v. infusions of lasmiditan

or sumatriptan in escalating cumulative doses, and carotid and coronary artery

diameters were measured.

Key Results:

Lasmiditan showed high selectivity for 5

‐HT

receptors. Moreover,

the functional potency of the analysed compounds to inhibit cAMP increase through

5

‐HT

receptor activation positively correlated with their potency to contract HCA.

In isolated human arteries, sumatriptan, but not lasmiditan, induced contractions.

Likewise, in vivo, sumatriptan decreased coronary and carotid artery diameters at

clinically relevant doses, while lasmiditan was devoid of vasoconstrictor activity at

all doses tested.

-This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

© 2019 The Authors. British Journal of Pharmacology published by John Wiley & Sons Ltd on behalf of British Pharmacological Society

Abbreviations: 5‐CT, 5‐carboxamidotryptamine; ANCOVA, analysis of covariance; DAP, diastolic arterial pressure; HCA, human coronary artery; LCX, left circumflex; MAP, mean arterial pressure; RANCOVA, repeated measure analysis of covariance; SAP, systolic arterial pressure

Funding information

CoLucid Pharmaceuticals Inc.; Consejo Nacional de Ciencia y Tecnología, Grant/ Award Numbers: 219707, 409865 and 410778; International Headache Society; Nederlandse Organisatie voor

Wetenschappelijk Onderzoek, Grant/Award Number: Vidi grant 917.113.349; Eli Lilly and Company

Conclusions and Implications:

Lasmiditan is a selective 5

‐HT

receptor agonist

devoid of vasoconstrictor activity. This may represent a cardiovascular safety

advan-tage when compared to the triptans.

1

I N T R O D U C T I O N

Migraine is a neurological disease characterized by throbbing unilateral headaches of moderate to severe intensity, accompanied by nausea, vomiting, photophobia, and/or phonophobia (Headache Classification Committee of the International Headache Society, 2018). It has an esti-mated prevalence of 15% in the global population, with women being three times more affected than men (Kurth et al., 2016; Vos et al., 2012). Currently, the specific therapies for acute antimigraine treatment are the triptans, selective 5‐HT1B/1D receptor agonists that also

display varying levels of5_‐HT1Freceptor affinity. Unfortunately, not all patients respond to triptans (Diener & Limmroth, 2001), and they are contraindicated in patients with cardiovascular disease, due to their contractile properties via activation of 5‐HT1Breceptors in coronary arteries (Chan et al., 2014; Dodick et al., 2004; Labruijere et al., 2015; MaassenVanDenBrink, Reekers, Bax, Ferrari, & Saxena, 1998). There-fore, there is a need for novel antimigraine drugs for the patients that do not respond to the current available treatments, but also, for those patients with cardiovascular disease.

Based on results from preclinical studies, the 5‐HT1Freceptor ago-nist,lasmiditan, was developed for acute antimigraine treatment (John-son et al., 1997; Nel(John-son et al., 2010) and Phase III trials showed positive results (Goadsby et al., 2019; Kuca, Silberstein, Wietecha, Berg, Dozier, & Lipton, 2018). Considering the increased cardiovascular risk of migraine patients (Buse, Reed, Fanning, Kurth, & Lipton, 2017; Kurth et al., 2016; Sacco & Kurth, 2014; Schurks et al., 2009), and the presence of 5_‐HT1F receptors in the vasculature (Bouchelet, Case, Olivier, & Hamel, 2000; Bouchelet, Cohen, Case, Séguéla, & Hamel, 1996; Nilsson et al., 1999), it is important to determine whether lasmiditan lacks affin-ity for human 5‐HT1B receptors and whether activation of 5‐HT1F receptors will result in vasoconstrictive responses. On this basis, the aim of this study was to investigate the pharmacological properties of lasmiditan and in particular (a) to assess the selectivity and functional activity of lasmiditan, triptans, and other 5‐HT receptor ligands on var-ious human 5_{‐HT receptors; (b) to analyse its potential to induce} vaso-constriction in in vitro (isolated human proximal and distal coronary, internal mammary, and middle meningeal arteries) and in vivo (carotid and coronary artery diameters in anesthetized dogs) preclinical models; and (c) to compare our findings with lasmiditan to those obtained with sumatriptan, one of the most prescribed triptans to acutely treat migraine.

We hypothesize that, unlike sumatriptan, lasmiditan selectively activates the human 5_‐HT1Freceptor and does not induce vasocon-striction in the above in vitro (including human coronary arteries) and

in vivo vascular models.

2

_{M A T E R I A L S A N D M E T H O D S}

2.1

Cell membrane preparation

CHO_{‐K1 cells (RRID:CVCL_0214) expressing the human recombinant} 5‐HT1A, 5‐HT1B,5‐HT1D,5‐ht1E, 5‐HT1F,5‐HT2A,5‐HT2B, or5‐HT7 receptors were grown prior to the test in media without antibiotic at Ogeda SA (Gosselies, Belgium). Cells were prepared using a protocol from Ogeda. In brief, cells were harvested by scraping from the culture vessels in ice‐cold Ca2+_{‐ and Mg}2+_{‐free PBS. The cells were then} centri-fuged for 10 min at 5,000×g and 4°C, and the pellets were suspended in buffer A (15‐mM Tris–HCl pH 7.5, 2‐mM MgCl2, 0.3‐mM EDTA, and 1‐ mM EGTA) and homogenized in a glass_{–glass homogenizer. The crude} membrane fraction was collected by two consecutive centrifugation steps at 35,000×g and 4°C for 30 min separated by a washing step in buffer A. The final membrane pellet was suspended in buffer B (75‐ mM Tris_{–HCl pH 7.5, 12.5‐mM MgCl}2, 0.3‐mM EDTA, 1‐mM EGTA, and 250‐mM sucrose) and flash‐frozen in liquid nitrogen. Protein con-tent was determined by the BCA method (Interchim, UP40840A).

2.2

Radioligand binding competition assay

Competition binding was performed in duplicate in the wells of a 96‐well plate containing binding buffer (optimized for each receptor), cell membrane extracts (approximately 20,000 cells distributed in the 96‐well plate), radiotracer, and test agonist. Non‐specific binding was determined by co_{‐incubation with 200‐fold excess of competitor.} Cells were incubated and exposed to varying concentrations (1 pM

What is already known

• Lasmiditan is effective in the acute treatment of migraine.

What this study adds

• Lasmiditan selectively activates the human 5‐HT1F receptor and is devoid of vasoconstrictor activity.

What is the clinical significance

• The lack of vasoconstriction may represent a safety advantage for migraine patients with cardiovascular disease.

to 10μM) of a range of displacer agonists (see below). The samples were incubated in a final volume of 0.1 ml and then filtered over Unifilter plates (Perkin Elmer, Massachusetts, USA) pretreated for 2 hr to limit tracer non_{‐specific binding. Filters were washed five} times with 0.5 ml of ice‐cold washing buffer (tris 50 mM pH 7.4) and 50_{μl of Microscint 20 (Perkin Elmer) were added to each filter.} The plates were incubated 15 min at room temperature on an orbital shaker and then counted with a TopCount_{™ (Perkin Elmer) for 1 min} per well.

2.2.1

_{cAMP HTRF assay for G}

_{coupled receptors}

Concentration–response curves were performed in parallel with the agonists. For agonist tests, 12_{μl of cells were mixed with 6 μl of the} test compound (at increasing concentrations) and 6 μl of forskolin and then incubated for 30 min at room temperature. After addition of the lysis buffer and 1‐hr incubation, cAMP concentrations were estimated according to the manufacturer specification with the HTRF kit (Cisbio International, Codelet, France). In brief, increasing concen-trations of agonists were added to stably transfected cells in buffer in an Optiplate (PerkinElmer Life Sciences, Massachusetts, USA). The plates were incubated, and cells were then lysed by the addition of HTRF reagents (cAMP‐Cryptate and anti‐cAMP‐d2 reagents) and diluted in lysis buffer, followed by incubation at room temperature. As 5‐HT1B receptors have been reported in naïve CHO‐K1 cells (George, Bungay, & Naylor, 1997), we also tested 5_‐ carboxamidotryptamine(5‐CT, the reference agonist for the 5‐HT1B receptor), in CHO_{‐K1 cells transfected with a non‐5‐HT, GPCR.}

2.2.2

cAMP HTRF assay for G

coupled receptors

Concentration_{–response curves were performed in parallel. For} ago-nist tests, 12μl of cells were mixed with 12 μl of the test compound at increasing concentrations and then incubated 30 min at room tem-perature. After addition of the lysis buffer and 60 min incubation, cAMP concentrations were estimated, according to the manufacturer specification, with the HTRF kit (Cisbio International, Codelet, France). Briefly, increasing concentrations of agonists were added to stably transfected cells in buffer in an Optiplate (PerkinElmer Life Sciences). The plates were incubated, and cells were then lysed by the addition of HTRF reagents (cAMP‐Cryptate and anti‐cAMP‐d2 reagents) and diluted in lysis buffer, followed by incubation at room temperature.

2.2.3

IPOne HTRF assay

The assay was performed on adherent cells. For agonist testing, the medium was removed and 20μl of assay buffer plus 20 μl of the stud-ied agonist or the reference agonist were added in each well. The plate was incubated for 60 min at 37°C with 5% CO2. IP1‐D2 reagent and anti_‐IP1cryptate reagents were dispensed in the wells, and IP1 con-centrations were then measured following the manufacturer instruc-tions (IPOne HTRF assay kit; Cisbio International, Codolet, France). In brief, increasing concentrations of agonists were added to stably

transfected cells in buffer in an Optiplate (PerkinElmer Life Sciences). The plates were incubated, and cells were then lysed by the addition of HTRF reagents (IP1‐D2 reagent and anti‐IP1 cryptate reagents) and diluted in lysis buffer, followed by incubation at room temperature.

2.2.4

GTP

S assay

For agonist testing, membrane extracts expressing the receptor of interest was mixed with GDP. In parallel, GTPγ[35S] was mixed with the beads just before starting the reaction. The following reagents were successively added in the wells of an Optiplate (Perkin Elmer): 50_{μl of reference ligand, 10 μl of assay buffer, 20 μl of the cells:} GDP mix, and 20μl of the GTPγ[35S]: beads mix. The plate was incu-bated for 60 min, then centrifuged and counted with a PerkinElmer TopCount™ reader.

2.2.5

_{Agonists tested}

5‐HT (serotonin), 5‐CT, ergotamine, sumatriptan, zolmitriptan, naratriptan, rizatriptan, almotriptan, eletriptan, frovatriptan, donitriptan, avitriptan, alniditan, lasmiditan, LY334370, and LY344864were tested. The radioligands and reference compounds used for the radioligand and second messenger studies are specified in Tables S1 and S2.

2.3

Human isolated arteries

2.3.1

_{Coronary arteries}

Coronary arteries were obtained from six“heart beating” organ donors (three males and three females; 48_{–62 years), who died of non‐cardiac} disorders less than 24 hr before the tissue was taken to the laboratory. The hearts were provided by the Heart Valve Bank Beverwijk Bank (nowadays ETB‐BISLIFE Tissue Bank) at that time still located in Rot-terdam, from Dutch post_{‐mortem donors, after donor mediation via} Bio Implant Services/Eurotransplant Foundation (Leiden, The Nether-lands), following removal of the aortic and pulmonary valves for homo-graft valve transplantation. All donors gave permission for research. Immediately after circulatory arrest, the hearts were stored at 4°C in a sterile organ protecting solution and were brought to the laboratory within the first 24 hr of death. After arrival at the laboratory, the right proximal (internal diameter 3–5 mm) and distal (internal diameter 0.5– 1 mm) portions of the coronary artery were dissected and placed in a cold, oxygenated (95% O2/5% CO2) Krebs buffer solution of the fol-lowing composition: 118_{‐mM NaCl, 4.7‐mM KCl, 2.5‐mM CaCl}2, 1.2‐ mM MgSO4, 1.2‐mM KH2PO4, 25‐mM NaHCO3, and 8.3‐mM glucose; pH 7.4.

2.3.2

Internal mammary arteries

Internal mammary arteries (internal diameter 2_{–3 mm) were obtained} perioperatively from 18 patients (16 males and two females; 51–

80 years) undergoing coronary bypass surgery. The tissue was immedi-ately placed in a sterile organ_{‐protecting solution and was brought to} the laboratory within 15 min. Subsequently, the artery was cleaned of connective tissue and placed in a cold, oxygenated Krebs buffer solution (for composition, see above).

2.3.3

_{Middle meningeal arteries}

Middle meningeal arteries (internal diameter 0.5–1.5 mm) were obtained from the dura mater of six patients (two males and four females; 12–68 years) who underwent neurosurgery. The dura mater, together with a small piece of the meningeal artery, was collected in a sterile organ‐protecting solution and immediately transported to the laboratory. The dura mater and connective tissue were dissected, and the artery was placed in a cold, oxygenated Krebs solution of the follow-ing composition: 119_{‐mM NaCl, 4.7‐mM KCl, 1.25‐mM CaCl}2, 1.2‐mM MgSO4, 1.2‐mM KH2PO4, 25‐mM NaHCO2, and 11.1‐mM glucose; pH 7.4.

All arteries were used on the same day or stored overnight and used the following day for functional experiments. The studies on cor-onary arteries were approved by the Scientific Advisory Board of the Rotterdam Heart Valve Bank. The Medical Ethics Committee of the Erasmus Medical Center, Rotterdam, approved the study protocols with regard to mammary arteries and middle meningeal arteries.

2.4

Isometric tension measurements

Proximal coronary arteries were cut into segments of 2_{‐ to 4‐mm} length, excluding distinct, macroscopically visible atherosclerotic lesions. The segments were mounted on stainless steel hooks in 15_‐ ml organ baths filled with oxygenated Krebs buffer solution at 37°C. After equilibration for at least 30 min and a wash every 15 min, the vessel segments were stretched to a stable tension of about 15 mN, with the optimal pre_{‐tension as determined earlier (Labruijere et al.,} 2015). Changes in tissue tension were measured with an isometric force transducer (Harvard, South Natick, MA, USA) and recorded on a flatbed recorder (Servogor 124, Goerz, Neudorf, Austria).

The distal coronary, internal mammary, and middle meningeal arteries were cut into circular 1‐ 2‐mm‐long segments and mounted in Mulvany myographs (Danish Myo Technology, Aarhus, Denmark) between two parallel small stainless‐steel wires (40‐μm calibre). All the baths were filled with warm Krebs buffer (37°C) and aerated with carbogen. The tension was normalized to 90% of l100for all segments, the diameter when transmural pressure equals 100 mm Hg (Mulvany & Halpern, 1977). Data of these vessels were recorded using a LabChart data acquisition system (AD Instruments Ltd, Oxford, UK).

2.4.1

_{Experimental protocols}

A paired parallel set up (i.e., all compounds were tested in different segments obtained from the same artery) was used. Initially, all seg-ments were exposed to 30‐mM KCl to “prime” the tissue for stable

contractions. After washout, the tissue was exposed to 100‐mM KCl to determine the maximal contractile response. After further washout, a concentration–response curve to vehicle, sumatriptan, or lasmiditan was constructed, using whole logarithmic steps from 1 nM up to 10 μM. After finishing the curve and washing several times until reaching equilibrium, the functional integrity of the endothelium was verified by observing relaxation to substance P (10 nM; coronary and meningeal arteries) orbradykinin(1_{μM; mammary arteries), after} precontraction with the TxA2analogueU46619(10–100 nM; Chan et al., 2014; Labruijere et al., 2015).

Furthermore, in the internal mammary arteries, a concentration– response curve to lasmiditan and sumatriptan was also constructed after adding threshold concentrations of U46619 (i.e., concentrations eliciting a contraction of ~10% of 100_{‐mM KCl response, determined} in quarter logarithmic steps), used to unmask contractile properties of some agonists in the presence of an increased tension, as previously described (MaassenVanDenBrink et al., 2000). These contractile responses were evaluated post hoc in the absence (relaxation to bra-dykinin <18%) or presence (relaxation to brabra-dykinin >18%) of func-tional endothelium; for this, endothelial function data was divided in percentiles, where values below the 50th percentile were considered without endothelium, and above the 50th percentile were considered with endothelium. Also, segments were preincubated with clinically relevant concentrations of sumatriptan (0.3_{μM) or lasmiditan (1 μM)} and followed by a concentration–response curve to lasmiditan or sumatriptan, respectively, to evaluate the possible interactions (i.e., augmented vasoconstriction) between agonists. The clinically relevant concentration of sumatriptan was calculated as previously described (Labruijere et al., 2015); in the case of lasmiditan, it was estimated based on the Cmax observed in humans following a 100‐mg dose (0.25μM; Kovalchin, Ghiglieri, Zanelli, Ings, & Mathers, 2016).

2.5

_{Correlation between binding (pK}

_{) and the}

contractile potency of lasmiditan and other triptans

We related previous (MaassenVanDenBrink et al., 1998; MaassenVanDenBrink, Reekers, Bax, & Saxena, 2000; Parsons et al., 1998; van den Broek et al., 2002) and current data obtained (see Sec-tion 3) to the potency of these compounds to contract the human iso-lated coronary artery. In case a compound failed to contract the human coronary artery, a fixed pEC50value of 5 was set (see Table S3). Our pKivalues used for this correlation are in agreement with those previously published in the literature (Table S4; Figure S1).

2.6

Animal preparations

Although in vitro experiments with human isolated arteries provide invaluable information on vasoconstrictive responses in specific vas-cular beds, to discard haemodynamic changes after systemic adminis-tration of novel experimental therapeutic compounds, an in vivo model is necessary. The beagle dog is a well_{‐accepted species that has been} in use for several years to predict human cardiovascular responses to

novel experimental therapeutic compounds (Cason, Verrier, London, Mangano, & Hickey, 1987). Therefore, a total of 18 adult female bea-gle dogs (Canis familiaris) were selected from the CorDynamics, Inc. animal colony. These animals were obtained from Marshall BioResources (North Rose, NY, USA). Upon receipt at the Biologic Resources Laboratory (BRL) of the University of Illinois_{‐Chicago, dogs} were examined by BRL veterinary personnel to ensure acceptable health status. Veterinary care was provided by the veterinarians and staff employed by the BRL. Dogs were acclimatized for at least 7 days prior to use and were pair_{‐housed in runs (meeting the size} require-ment set forth by the USDA Animal Welfare Act) with various cage‐ enrichment devices. Room temperature set at 18_{–27°C, humidity at} 30–70%, and fluorescent lights timed to give a 12 hr‐light and 12 hr_{‐dark cycle. Harlan Certified Canine food (25% Protein Diet} #2025C, Harlan Teklad, Madison, WI, USA) was fed daily (500 g·day−1), and water was freely available in their runs. At the day of their termi-nal experiment, the animals were 10.0 to 11.5 months old, and their body weights ranged from 5.4 to 7.9 kg. Body weights were measured twice (approximately 1 week between measurements) prior to each animal's terminal procedure. Dogs were fasted for 16_{–18 hr prior to} dosing.

All experimental protocols were approved and conducted by CorDynamics in compliance with the U.S. FDA Good Laboratory Practice guidelines (21 CFR Part 58), the Animal Welfare Act, the Guide for the Care and Use of Laboratory Animals, and the Office of Laboratory Animal Welfare. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny, Browne, Cuthill, Emerson, & Altman, 2010) and with the recommendations made by the British Journal of Pharmacology (Curtis et al., 2018) and the edito-rial on reporting animal studies (McGrath & Lilley, 2015). For more specific details on design and statistical analysis, see Sections 2.7 and 2.8.

2.6.1

General methods

Dogs were dosed with morphine s.c. (1 mg·kg−1) approximately 10_– 20 min prior to administration of propofol anaesthesia i.v. (5– 6 mg·kg−1) to allow tracheal intubation. They were placed on a venti-lator with isoflurane delivered at 1–2% in oxygen to maintain anaes-thesia throughout the experiment, and s.c. morphine (0.5 mg·kg−1) was administered approximately every 2 hr while under anaesthesia. The local anaesthetic bupivacaine was infiltrated into the incision sites. A continuous 0.9% NaCl solution for injection drip (approxi-mately 10 ml·kg−1·hr−1) was maintained until the start of dosing at which time it was discontinued. Dogs were placed on a heating pad set to maintain the animal's body temperature at approximately 37° C, and their body temperature was monitored throughout the experi-ment by placing a rectal temperature probe. Additionally, surface ECG leads were placed for anaesthesia monitoring throughout the experi-mental protocol.

A mid‐lateral neck incision was made and the left common carotid artery was exposed. A Transonic Systems Inc. (Ithaca, NY, USA) blood flow probe and two Sonometric Corporation (London, Ontario,

Canada) crystals for arterial dimensional analysis were affixed to the artery. A left lateral thoracotomy (sixth, intercostal space) was per-formed, and the left circumflex (LCX) coronary artery was exposed. A Transonic Systems Inc. blood flow probe and two Sonometric Cor-poration crystals for arterial dimensional analysis were affixed to the artery. A solid_{‐state high‐fidelity pressure catheter (Millar Inc.,} Hous-ton, TX, USA) for measurement of arterial pressure (mean, MAP; sys-tolic, SAP; and diassys-tolic, DAP) and heart rate was inserted into a femoral artery and secured in place with silk suture. An indwelling catheter was placed into the femoral vein for collecting blood (2 ml) prior to the start of dosing and at the end of each 20 min infusion period (see experimental protocol) for bioanalytical analysis.

2.6.2

_{Experimental protocol}

Upon completion of the general instrumentation, a 15‐min equilibrium period was allowed for a stable haemodynamic condition. Baseline values (defined as the average of the three 5‐min values at the afore-mentioned 15 min) of MAP, heart rate, and carotid and LCX coronary artery diameter and flow were determined. Subsequently, the 18 dogs were randomly assigned into three groups (n = 6 each), which received vehicle (saline), lasmiditan (0.03, 0.13, 1.13, 4.13, and 11.13 mg·kg−1), or sumatriptan (0.03–11.3 mg·kg−1) respectively. All treatments were filtered through a 0.2‐μM membrane and administered i.v. in the esca-lating cumulative doses mentioned above. Dose intervals amongst dif-ferent treatments were administered each 20 min. At the end of the experiment, dogs were killed while under anaesthesia via a barbiturate overdose.

2.7

_{Sample size calculation, randomization, and}

blinding

2.7.1

Sample size calculation

The animal sample size (n = 6 each group) was calculated by CorDynamics based on their previous studies (Cushing et al., 2009). For the in vitro studies, we based the experimental number (n = 5–7 each group) on previous studies from our group (Chan, Baun, et al., 2011; Labruijere et al., 2015).

2.7.2

_{Randomization}

For the in vitro experiments, all artery segments were cut into rings and randomly assigned to a bath, and then the treatment group was designed by using a table of random numbers.

For the in vivo experiments, the animals initially divided in sets (n = 6 each group as described above) were randomly assigned to study groups by CorDynamics staff.

2.7.3

Blinding

For the radioligand and second messenger assays, the analyst was not blinded to the compounds but to the research hypothesis. For

the vascular in vitro experiments, values were calculated using the dose_{–response auto‐analyse selection feature of LabChart. During} the analysis, the investigator was unaware of which concentration response curve was being analysed. The in vivo experimental values (i.e., the changes in MAP or artery diameter) in each group of ani-mals were simultaneously obtained by at least two different CorDynamics investigators, with at least one of the investigators blinded.

2.8

_{Data presentation and statistical analysis}

Pharmacology on experimental design and analysis in pharmacology

(Curtis et al., 2018).

2.8.1

_{Radioligand binding and second messenger}

activity

For binding competition and second messenger activity assays, refer-ence compounds were tested at several concentrations in duplicate to obtain a concentration_{–response curve, and an estimated pEC}50 (negative logarithm of the concentration eliciting 50% of the maximal contractile response, i.e., Emax) or pIC50 value (negative logarithm of the concentration that displaced 50% of the radioligand) was calcu-lated using XLFit (IDBS, Guildford, UK). Additionally, the reference values obtained were compared to historical values obtained from the same receptor and used to validate the experimental session. A session was considered as valid only if the reference value was found to be within a 0.5 log interval from the historical value, for assays where historical values (determined in at least five experiments) were available (Abourashed, Koetter, & Brattström, 2004; Barac et al., 2012; Sun, Blanton, Gabriel, & Canney, 2005). For the new assays developed in this study (i.e., 5_‐ht1Ereceptor), the two independent pIC50 deter-mined must be concordant with a 1 log unit interval for the assays to be validated. When less than 50% inhibition of binding or second messenger activation was obtained at 10μM, a pIC50/pEC50of“<5” was set

2.8.2

_{Experiments with human isolated arteries}

For the in vitro studies using human blood vessels, concentration– response curves were analysed using GraphPad software (GraphPad Software Inc., San Diego, CA, USA; RRID:SCR_002798) to determine pEC50values as previously reported (Labruijere et al., 2015). When a plateau in the concentration–response curve was not reached, the response observed with the highest concentration used (i.e., 100μM) was considered as Emax. Differences between pEC50and Emax values of the compounds were evaluated with Tukey's test, once an ANOVA for paired data had revealed that the samples represented

different populations. Values of P < .05 were considered to indicate significant differences.

2.8.3

In vivo studies

In the in vivo studies, each haemodynamic parameter was analysed with a repeated measure analysis of covariance (RANCOVA) for changes from baseline at time intervals of 5, 10, 15, and 20 min for each of the five dose levels. The model factored the treatment (TRT), the time after dose (TIME), and the interaction of time after dose with treatment (TRT * TIME). The SAS® procedure PROC MIXED was used for analysis with TIME as the repeated effect and ANIMAL as the subject. The covariance between errors from the same animal at different time points was selected based on the corrected Akaike's information criterion from selected covariance structures of VC, AR(1), UN, and CS. Non‐monotonic dose–responses were evalu-ated. Within the framework of the RANCOVA, comparisons were made for vehicle‐ versus lasmiditan‐treated animals and for vehicle‐ versus sumatriptan_{‐treated animals. If TRT * TIME was significant,} the comparisons were conducted for each time interval using an anal-ysis of covariance (ANCOVA) model with an effect for treatment and baseline as a covariate. If only the TRT effect was significant, the com-parison was conducted across the pooled time intervals for the overall phase only. These non‐monotonic treatment group comparisons were conducted at the P = .01 significance level. Baseline data were analysed with an ANOVA for each time interval. Factors in the model included treatment (TRT). All statistical analyses were conducted with SAS® version 9.2 (RRID:SCR_008567). After the database lock, post hoc analyses for coronary artery diameter and carotid artery diameter (primary endpoints) at the clinically relevant time interval 20 min (com-pletion of dose administration) for each cumulated dose (0.03_– 11.13 mg·kg−1) were performed for comparisons between sumatriptan and vehicle. A significance level of P _{≤ .025 was used for the} RANCOVA using Bonferroni correction of two tests (coronary and carotid artery diameter).

2.9

Materials

The compounds used in the present study (obtained from the sources indicated) were 5‐HT hydrochloride, naratriptan hydrochloride, almotriptan malate, avitriptan fumarate, and sumatriptan succinate (Sigma Chemical Co., St. Louis, MO, USA); lasmiditan hemisuccinate (Eli Lilly & Co., Indianapolis, IN, USA); 5_{‐CT maleate, sumatriptan} suc-cinate, zolmitriptan, rizatriptan benzoate, eletriptan hydrobromide, donitriptan hydrochloride, LY334370 hydrochloride, and LY344864 hydrochloride (Tocris Bioscience Co., Park Ellisville, MO, USA); ergot-amine tartrate (TEVA Pharmaceutical Industries Ltd., Petach Tivka, Israel); and alniditan salt (kind gift of Janssen Pharmaceutica, Beerse, Belgium).

All compounds were dissolved in distilled water or physiological saline for the in vitro and in vivo studies respectively. These vehicles had no effect on the baseline MAP values or artery diameter (not

shown). Fresh solutions were prepared for each experiment. The doses mentioned in this text refer to the free base of substances.

2.10

Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corre-sponding entries in http://www.guidetopharmacology.org, the com-mon portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2017/18 (Alexander et al., 2017).

3

_{R E S U L T S}

3.1

Pharmacological characterization of lasmiditan

As shown in Table 1, radioligand studies revealed that lasmiditan selec-tively binds to the human 5‐HT1Freceptor. On the other hand, the triptans almotriptan, avitriptan, eletriptan, frovatriptan, naratriptan, sumatriptan, and zolmitriptan showed affinity also for 5‐HT1B, 5‐ HT1D, and 5‐HT1Freceptors, while alniditan, donitriptan, ergotamine, and rizatriptan had affinity for 5‐HT1Band 5‐HT1Dreceptors. Most importantly, when analysing their second messenger activity, we observed that ergotamine is an agonist of the 5‐HT1A/B/D, 5‐HT2A/B, and 5_‐HT7receptors (but not of 5‐HT1Freceptors). Similarly, sumatrip-tan, zolmitripsumatrip-tan, naratripsumatrip-tan, almotripsumatrip-tan, eletripsumatrip-tan, frovatripsumatrip-tan, and avitriptan are agonists of the 5_‐HT1B/Dreceptors and also of the 5‐ HT1F receptor. Lasmiditan, as well as LY344864, displayed a high potency only for the 5_‐HT1F receptor (Table 2). In Figure 1, the

agonistic profile of the different antimigraine drugs tested on the 5‐ HT1B/D/Freceptors (i.e., relevant for migraine therapy) are represented. When comparing our results with those previously published, our values are in agreement with those in the literature (see Tables S4_–S5; Figure S1). Moreover, no functional responses to 5‐CT were observed in the CHO_{‐K1 cells transfected with an unrelated GPCR (Figure S2).}

3.2

_{Human isolated arteries}

In the human isolated coronary arteries, sumatriptan induced signifi-cant contractions in a concentration_{‐dependent manner in the} proxi-mal (Emax 39 ± 12%, pEC50 6.4 ± 0.2; n = 6) and distal (Emax 59 ± 41%, pEC506.02 ± 0.2; n = 6) coronary segments, even at clini-cally relevant concentrations (see Figure 2). In contrast, as compared with vehicle, lasmiditan was devoid of any significant contractile effects in both coronary segments, even at its highest concentration (100_{μM, supratherapeutic; E}max1 ± 1%).

Moreover, Figure 3 shows the effects of lasmiditan and sumatrip-tan on internal mammary arteries in the absence and presence of endothelium. Sumatriptan induced concentration‐dependent contrac-tions in blood vessels with (Emax 46 ± 18%, pEC50 6.07 ± 0.07;

n = 5) and without functional endothelium (Emax31 ± 12%, pEC50 5.6 ± 0.94; n = 7). After precontraction with threshold concentrations of U46619, sumatriptan also produced concentration‐dependent con-tractions in the presence (Emax63 ± 19%, pEC506.83 ± 0.05; n = 5) and absence (Emax59 ± 16%, pEC506.02 ± 0.59; n = 7) of functional endo-thelium. In marked contrast, vehicle and lasmiditan were devoid of any significant contractile effects in internal mammary arteries (Emax 1 ± 1% without endothelium; Emax0 ± 0% with endothelium), even TABLE 1 Summary of pIC50(negative logarithm of the molar concentration of these compounds at which 50% of the radioligand is displaced) and pKi(negative logarithm of the molar concentration of the Ki) values of individual antimigraine drugs at 5‐HT receptors

Agonist

5_‐HT1A 5‐HT1B 5‐HT1D 5‐ht1E 5‐HT1F 5‐HT2A 5‐HT2B 5‐HT7

pIC50 pKi pIC50 pKi pIC50 pKi pIC50 pKi pIC50 pKi pIC50 pKi pIC50 pKi pIC50 pKi

Ergotamine tartrate 9.19 9.70 8.87 9.34 8.63 9.31 6.08 6.39 6.71 7.13 7.62 8.14 7.73 7.94 7.13 7.23 Sumatriptan succinate 6.63 7.14 7.81 8.29 8.31 9.00 5.42 5.72 7.13 7.55 <5 <5 <5 <5 6.10 6.19 Zolmitriptan 7.28 7.79 8.85 9.33 9.28 9.97 7.51 7.81 7.13 7.55 <5 <5 <5 <5 6.97 7.06 Naratriptan hydrochloride 7.31 7.82 8.75 9.22 8.62 9.30 7.83 8.13 8.33 8.75 <5 <5 <5 5.08 5.84 5.93 Rizatriptan benzoate 6.81 7.32 7.51 7.99 8.15 8.83 6.48 6.78 6.4 6.82 <5 <5 5.30 5.51 <5 <5 Almotriptan malate 6.23 6.73 7.97 8.45 7.57 8.26 <5 <5 7.15 7.57 <5 <5 <5 <5 6.36 6.46 Eletriptan hydrobromide 8.20 8.71 8.80 9.28 9.31 9.99 6.91 7.21 7.35 7.77 5.42 5.94 6.14 6.35 6.61 6.70 Frovatriptan racemate 6.83 7.34 8.09 8.57 8.10 8.78 <5 5.18 6.50 6.92 <5 <5 <5 <5 6.88 6.97 Donitriptan hydrochloride 7.42 7.93 9.29 9.77 9.18 9.86 5.47 5.77 <5 5.18 5.83 6.35 5.88 6.09 6.12 6.21 Avitriptan fumarate 7.20 7.71 8.32 8.80 8.42 9.11 5.15 5.45 6.69 7.11 5.11 5.63 5.73 5.94 6.03 6.12 Alniditan dihydrochloride 8.81 9.32 8.93 9.41 8.66 9.35 5.98 6.28 6.02 6.44 <5 5.43 6.67 6.88 7.16 7.26 Lasmiditan hemisuccinate 5.88 6.39 5.54 6.02 5.62 6.31 5.54 5.84 8.09 8.51 <5 <5 5.01 5.22 <5 <5 LY334370 hydrochloride 7.98 8.49 6.74 7.21 6.24 6.92 6.83 7.13 9.03 9.45 5.11 5.63 5.98 6.19 5.66 5.75 LY344864 hydrochloride 6.12 6.63 6.13 6.61 5.83 6.52 6.05 6.35 8.38 8.80 5.11 5.63 5.31 5.52 5.69 5.78

Note. The lesser than symbol (<) indicates that less than 50% inhibition of binding was obtained at 10μM. The radioligands used and their concentrations

TABLE 2 Summary of pEC50values of cAMP (5‐HT1A/B/E/Fand 5‐HT7), GTPγS (5‐HT1A/B/D/E/F), and IP (5‐HT2) assays of individual antimigraine drugs at 5‐HT receptors

Agonist

5_‐HT1A 5‐HT1B 5‐HT1D 5‐ht1E 5‐HT1F _5‐HT

2A 5‐HT2B 5‐HT7

cAMP GTP_γS cAMP GTP_γS GTP_γS cAMP GTP_γS cAMP GTP_γS IP IP cAMP

Ergotamine tartrate 9.78 9.63 9.94 9.52 9.43 5.95 5.74 5.97 6.30 9.25 8.72 7.09 Sumatriptan succinate <5 <5 7.32 7.91 8.30 5.99 5.79 8.03 6.80 <5 <5 5.22 Zolmitriptan <5 5.52 7.87 8.42 9.51 8.18 7.81 8.00 6.67 <5 <5 6.28 Naratriptan hydrochloride <5 6.52 8.05 8.86 8.80 7.75 8.17 8.38 8.05 <5 <5 <5 Rizatriptan benzoate <5 <5 7.08 7.56 8.11 7.34 6.90 6.54 5.91 <5 5.49 <5 Almotriptan malate <5 5.48 7.08 7.85 7.75 <5 <5 7.79 6.90 <5 5.20 <5 Eletriptan hydrobromide 5.74 6.38 8.00 8.09 9.04 7.53 6.90 8.13 6.88 6.07 6.81 6.45 Frovatriptan racemate <5 6.12 7.98 8.14 8.36 5.04 <5 7.10 6.35 <5 <5 7.42 Donitriptan hydrochloride 5.94 6.74 9.96 9.52 9.51 <5 <5 <5 <5 8.10 7.61 5.23 Avitriptan fumarate <5 6.19 8.57 8.68 9.27 5.52 <5 7.09 6.05 6.91 6.41 5.38 Alniditan dihydrochloride 7.00 7.29 8.87 8.90 8.20 5.68 5.21 5.92 5.17 <5 7.15 6.32 Lasmiditan hemisuccinate <5 <5 <5 <5 6.64 6.17 5.34 8.43 7.80 <5 <5 <5 LY334370 hydrochloride 5.84 6.96 6.52 5.80 6.92 7.53 6.95 9.08 9.38 <5 <5 <5 LY344864 hydrochloride <5 <5 <5 5.82 6.93 6.22 6.12 8.72 7.85 <5 <5 <5

Note. These values represent the negative logarithm of the molar concentration of these compounds at which 50% of their maximal response is exerted.

The lesser than symbol (<) indicates that less than 50% response was obtained at 10_{μM. The reference compounds used and their concentrations are} described in Table S2.

FIGURE 1 Summary of the agonist profiles (pEC50> 7) of the antimigraine drugs tested on the 5‐HT1B, 5‐HT1D, and 5‐HT1Freceptors. Redrawn from Rubio‐Beltran, Labastida‐Ramirez, Villalon, and MaassenVanDenBrink (2018)

after a modest precontraction with U46619 (Emax 0 ± 1% without endothelium; Emax1 ± 1% with endothelium).

In the middle meningeal artery, sumatriptan induced significant concentration_{‐dependent contractions (E}max 73 ± 13%, pEC50 6.32 ± 0.15; n = 6), whereas lasmiditan did not induce any significant contraction at all concentrations tested (Emax0 ± 0%, Figure 4).

4

I N T E R A C T I O N E X P E R I M E N T S

As shown in Figure 5, in the internal mammary arteries, after preincubation with lasmiditan (1μM), no changes in the contractile responses to sumatriptan were observed when compared to the concentration–response curve to sumatriptan alone (Emax59 ± 16%, pEC505.34 ± 0.1 vs. Emax51 ± 19%, pEC505.71 ± 0.7; n = 5 each). In addition, the highest concentration of lasmiditan produced a non‐ significant vasodilation when preincubated with sumatriptan's clini-cally relevant concentration (0.3 μM) when compared to the concentration_{–response curve to lasmiditan without sumatriptan} (Emax− 4.8 ± 5.95% vs. Emax0 ± 0%; n = 5 each).

4.1

_{Correlation between binding (pK}

_{) and the}

contractile potency of lasmiditan and other triptans

As shown in Figure 6, the potency of the compounds tested to con-tract the human isolated coronary artery was positively correlated with their potency to bind the 5_‐HT1Breceptor, whereas it was nega-tively correlated for the 5‐HT1F receptor. This was also observed when correlating the pEC50values obtained in our second messenger assays and the contractile potency of the compounds tested in the human coronary arteries (Figure S3).

4.2

In vivo studies

In anaesthetised dogs, a directly proportional relationship was observed between lasmiditan and sumatriptan cumulative i.v. doses and their plasma concentrations; these latter values were used for val-idating the concentrations used in the in vitro studies (data not shown). Moreover, as shown in Figure 7, changes in carotid artery diameter were not statistically significant in the lasmiditan‐treated group as FIGURE 2 Contractile responses to lasmiditan and sumatriptan (1 nM–100 μM) in the isolated human proximal (left) and distal (right) coronary arteries; *P < .05, significantly different as indicated; n = 6 each

FIGURE 3 Contractile responses to sumatriptan and lasmiditan (1 nM–100 μM) in the absence (left) and presence (right) of a threshold precontraction with U46619 (1–10 nM) in the isolated human internal mammary arteries with (upper panel, n = 5) and without (lower panel,

compared to the time‐matched vehicle control group. In contrast, as expected, sumatriptan induced dose_{‐dependent decreases in carotid} artery diameter, although these effects were statistically significant only at the doses of 0.13 mg·kg−1 (clinically relevant) and 11.13 mg·kg−1(Figure 7). In the LCX coronary artery, lasmiditan failed to induce any statistically significant change in diameter at any dose. Conversely, statistically significant decreases in the LCX coronary artery diameter were observed at all doses in sumatriptan_‐treated ani-mals as compared to the time‐matched vehicle control animals, even at the lowest dose of 0.03 mg·kg−1, which already corresponds to a clinically relevant dose.

Carotid blood flow was not significantly different after vehicle or clinically relevant doses of lasmiditan (0.03–1.13 mg·kg−1). Lasmiditan decreased carotid blood flow significantly, but only after the supratherapeutic cumulative doses of 4.13 and 11.13 mg·kg−1. In contrast, sumatriptan elicited a statistically significant rapid, dose_‐ dependent, decrease in carotid blood flow at all doses tested.

Regarding coronary blood flow, the administration of vehicle, lasmiditan, or sumatriptan did not elicit any statistically significant changes (data not shown).

Heart rate was stable over the course of the study, and no significant changes were observed in the lasmiditan or vehicle groups. In the sumatriptan_{‐treated group, cumulative doses of 4.13 and} 11.13 mg·kg−1elicited dose‐dependent decreases in heart rate, which were statistically significant, with a peak decrease at 90 min of 16.5 ± 6 bpm (data not shown). MAP, SAP, and DAP showed no signif-icant changes in either sumatriptan or lasmiditan_{‐treated groups as} compared to the time‐matched vehicle group at cumulative doses of up to 4.13 mg·kg−1. At higher doses, both lasmiditan_{‐ and} sumatriptan‐treated groups showed a dose‐dependent trend to decrease MAP, SAP, and DAP; however, these changes were not sta-tistically significant (Figure 7).

5

D I S C U S S I O N

The current study was designed to investigate the selectivity and vasoconstrictor profile of lasmiditan, which belongs to a novel class of acute antimigraine drugs, the ditans. According to its binding and functional activity, it was confirmed that lasmiditan is a highly selec-tive agonist of the 5_‐HT1F receptor. Moreover, as lasmiditan was developed based on the premise that coronary vasoconstriction is a side effect of the triptans attributed to 5_‐HT1Breceptors, we studied the vasoconstrictor potential of 5‐HT1Fagonism in two different vas-cular models and compared our in vitro and in vivo results to those obtained with sumatriptan, since this is the “gold standard” triptan for acute antimigraine treatment. This allowed us to compare the results from the current study with results obtained earlier. In accor-dance with our previous work (Chan et al., 2014), sumatriptan induced a concentration‐dependent contraction in human isolated coronary arteries, which tended to be larger in distal than in proximal coronary artery segments. This contraction was apparent at clinically relevant concentrations and is most likely due to activation of 5_‐HT1Breceptors in vascular smooth muscle (Chan et al., 2014). In contrast, lasmiditan FIGURE 4 Contractile responses to sumatriptan and lasmiditan

(1 nM–100 μM) in the isolated human middle meningeal arteries; *P < .05, significantly different as indicated; n = 6 each

FIGURE 5 Contractile responses to sumatriptan and lasmiditan (1 nM–100 μM) in the internal mammary artery, after preincubation with the clinically relevant concentration of sumatriptan (0.3_{μM) or lasmiditan (1 μM), and followed by a concentration–response curve to lasmiditan or} sumatriptan, respectively (n = 6 each)

did not induce a contraction at concentrations up to 100_{μM (≥100×} the clinically relevant concentrations) in either proximal or distal coro-nary arteries. Although moderate to intense expression of mRNA encoding the 5‐HT1Freceptor in human coronary arteries has been described (Nilsson et al., 1999), the presence of mRNA does not neces-sarily mean protein expression, which may well be the case. Thus, the physiological role of this receptor in blood vessels remains to be determined.

Subsequently, we performed more in_{‐depth experiments in the} internal mammary artery, where we studied the influence of endothe-lial functional quality, and the effects of a precontraction induced by the TxA2 analogue U46619, as such a precontraction is known to “unmask” or augment contractions to other ligands, such as sumatrip-tan (MaassenVanDenBrink, van den Broek, de Vries, Upton, et al., 2000). As observed in the coronary artery, sumatriptan contracted the internal mammary artery, similarly in both segments with and without functionally active endothelium. In accordance with earlier observations (MaassenVanDenBrink, van den Broek, de Vries, Upton, et al., 2000), the contractions to sumatriptan were augmented in the presence of U46619. In contrast, lasmiditan did not induce any con-traction in the absence or presence of U46619 in either vessel seg-ments with or without endothelium. Interestingly, in the rabbit saphenous vein, precontraction withPGF2α unmasked a contractile

response to the 5‐HT1Freceptor agonists, LY334370 and LY344864, but only after high concentrations (>10_{μM), and therefore likely due} to activation of vascular 5‐HT1Breceptors (Cohen & Schenck, 2000).

Hence, the absence of contractile responses with high concentrations of lasmiditan, even in precontracted vessels, is surprising, given the difference in affinity between sumatriptan and lasmiditan to the 5_‐ HT1Breceptor. However, binding affinity does not always correlate with second messenger activation and biological response (Colquhoun, 1998). Therefore, while our radioligand studies (Table 1) showed a ~100_{‐fold binding difference to the 5‐HT}1Breceptor between suma-triptan (pIC50= 7.81) and lasmiditan (pIC50= 5.54), our cAMP assays (Table 2) showed that the functional potency (pEC50) of sumatriptan was 7.32 and lasmiditan was <5. As we could not determine the pre-cise pEC50value of lasmiditan, the potency difference between both compounds could be larger than 1,000‐fold and thus explain the com-plete absence of vasoconstrictive responses even at supra_{‐therapeutic} concentrations such as 100μM. This could represent a cardiovascular safety advantage over its triptan predecessors.

Additionally, as contraction of the meningeal artery is thought to contribute to the antimigraine effects of the triptans (Benemei et al., 2017; Chan, Vermeersch, de Hoon, Villalón, & MaassenVanDenBrink, 2011; Rubio_{‐Beltran et al., 2018), but is not a class effect of all} antimigraine drugs (e.g., gepants), we investigated the contractions to sumatriptan and lasmiditan in human meningeal arteries. In accordance to the previously described craniovascular selectivity of the triptans (MaassenVanDenBrink et al., 2000; van den Broek et al., 2002), con-tractions to sumatriptan were larger in this dural artery than those in the proximal coronary artery. However, as we have also previously shown, contractions to sumatriptan in distal coronary artery (and FIGURE 6 Correlation between the pKi values obtained in our study and the contractile potency of lasmiditan, triptans (sumatriptan, zolmitriptan, naratriptan, rizatriptan, eletriptan, frovatriptan, donitriptan, and avitriptan), and other 5_{‐HT receptors ligands (ergotamine, alniditan,} 5HT, and 5‐carboxamidotryptamine) in human isolated coronary arteries; N.S., non‐significant;*P < .05, significant correlation

internal mammary artery) were not significantly different from those in meningeal artery (Chan et al., 2014). In contrast, lasmiditan was devoid of vascular effects in this cranial vessel. Therefore, the efficacy of lasmiditan as acute migraine treatment may be due to inhi-bition of CGRP release from perivascular fibres or direct central (antinociceptive) modulation (Rubio_{‐Beltran et al., 2018).}

Our binding studies showed that, as mentioned previously, most of the triptans available in the market, namely, almotriptan, frovatriptan, naratriptan, sumatriptan, and zolmitriptan, are also agonists of the 5‐ HT1Freceptor (Figure 1). Furthermore, the correlation between bind-ing and the contractile potency of the compounds tested revealed that the potency of the agonists to contract the HCA positively correlated to their potency to bind the 5‐HT1Breceptor, whereas it negatively correlated for the 5‐HT1F receptor (Figure 6), and this was also observed when correlating the contractile potency and second

messenger activation (Figure S3). Moreover, when analysing the corre-lation between second messenger activation by 5‐HT1Bversus 5‐HT1F receptors, also a negative correlation was observed (Figure S4). These results, together with our in vitro data, suggest that although the mRNA of both receptor subtypes has been described in human vascu-lature (Bouchelet et al., 1996; Bouchelet et al., 2000; Chan et al., 2009; Chan et al., 2014; Nilsson et al., 1999; Parsons et al., 1998; van den Broek et al., 2002), only activation of the 5‐HT1Breceptor will result in vasoconstriction of, at least, coronary, mammary, and meningeal arteries, whereas activation of the 5_{‐ HT}1Freceptor will not. This could suggest that either 5‐HT1Freceptors in human vasculature are not functional or that 5‐HT1Freceptor mRNA is not translated to protein. When considering the acute haemodynamic effect of sumatriptan in humans, it is well known that after subcutaneous administration, there are vasopressor responses in the systemic arterial circulation

FIGURE 7 Changes in the left circumflex (LCX) coronary artery diameter (a), carotid artery diameter (b), and mean arterial blood pressure (c) after the continuous infusion of lasmiditan and sumatriptan (0.03–

11.13 mg·kg−1each) or the corresponding infusion volumes of vehicle in female beagle dogs (n = 6 each). *P < .05, significantly different from vehicle; post hoc analysis (see methods)

and coronary artery vasoconstriction (MacIntyre, Bhargava, Hogg, Gemmill, & Hillis, 1993). Although we observed carotid and coronary vasoconstriction in anaesthetized dogs, there were no significant increases in blood pressure, as previously reported in this animal model (Villalón & Terrón, 1994). In fact, after high doses of sumatrip-tan, a non_{‐significant tendency to decrease blood pressure and} signif-icant decreases in heart rate were observed, most probably due to inhibition of vascular and cardiac sympathetic outflows via the stimu-lation of prejunctional 5‐HT1B/1Dreceptors on perivascular (Villalón et al., 1998) and cardiac (Sánchez_{‐López et al., 2003; Sánchez‐López} et al., 2004) sympathetic nerves. Lasmiditan only showed a trend to decrease blood pressure at the highest (supratherapeutic dose), which, based on the affinity of lasmiditan (see Table 1), could be due to a non‐ selective activation of prejunctional 5_‐HT1Dreceptors and subsequent inhibition of sympathetic perivascular nerves (Villalón et al., 1998). Admittedly, this has not been shown directly in dogs but in pithed Wistar rats, and, in patients, no changes in blood pressure have been observed (Farkkila et al., 2012; Ferrari et al., 2010). Further experi-ments, falling beyond the scope of the present study, would be required to shed more light on the mechanisms behind these responses, which are only observed at non‐clinically relevant doses.

In summary, our in vitro and in vivo results indicate that lasmiditan is devoid of contractile properties in isolated human and anaesthetized dog arteries respectively. This might be of particular relevance in migraine patients who have a high risk of developing cardiovascular disease, such as subjects with hemiplegic migraine, prolonged migraine with aura, or with established cardiovascular disease. Clearly, further studies are needed to evaluate the safety of lasmiditan in these spe-cific patient populations and its effectiveness compared with triptans. Finally, clinical trials have shown that lasmiditan is effective for migraine treatment (Goadsby et al., 2019; Kuca et al., 2018), suggest-ing a mechanism of action (partly) different to that of the triptans (Rubio‐Beltran et al., 2018).

In conclusion, our data support our initial hypothesis that lasmiditan is a high‐affinity and highly selective agonist for the human 5_‐HT1Freceptor that is devoid of contractile properties in human iso-lated blood vessels and in anaesthetized canines.

A C K N O W L E D G E M E N T S

Dr Antoinette MaassenVanDenBrink was supported by the Nether-lands Organization for Scientific Research (Vidi Grant 917.113.349), whereas Prof. Carlos M. Villalón, Alejandro Labastida_{‐Ramírez, and} Eloísa Rubio‐Beltrán were supported by Consejo Nacional de Ciencia y Tecnología (CONACyT; Grant 219707 to C.M.V. as well as fellow-ships 410778 to A.L.R. and 409865 to E.R.B.; Mexico City). Dr Kristian A. Haanes was supported by a fellowship from the International Head-ache Society. This study was financially supported by a research grant from CoLucid Pharmaceuticals and Eli Lilly and Company.

A U T H O R C O N T R I B U T I O N S

E.R.B., E.Z., L.M., A.H.J.D., M.R.G., P.B.S., K.W.J., J.K., C.M.V., and A.M. v.d.B. contributed to conception and design. E.R.B., A.L.R., K.A.H., L. M., M.R.G., and P.B.S. performed the acquisition, analysis, and

interpretation of data. E.R.B. and A.M.v.d.B. drafted the manuscript. E.R.B., A.L.R., K.A.H., A.B., A.J.J.C.B., E.Z., L.M., A.H.J.D., M.R.G., P.B.S., K.W.J., J.K., C.M.V., and A.M.v.d.B. revised the manuscript. All authors approved the final version of the manuscript.

C O N F L I C T O F I N T E R E S T

E.R.B. and A.L.R. received travel support from Eli Lilly/CoLucid. E.Z. and C.M.V. received consultation fees from Eli Lilly/CoLucid. L.M., M.R.G. and P.B.S. performed experiments under a research contract with Eli Lilly/CoLucid. J.K. is former employee of Eli Lilly/CoLucid. K. W.J. is employee of Eli Lilly. A.M.v.d.B. received research grants and/or consultation fees from Amgen/Novartis, Eli Lilly/CoLucid, Teva, and ATI. All other authors declare no conflicts of interest.

D E C L A R A T I O N O F T R A N S P A R E N C Y A N D S C I E N T I F I C R I G O U R

This Declaration acknowledges that this paper adheres to the princi-ples for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines forDesign & Analysis, and

Animal Experimentation, and as recommended by funding agencies, publishers and other organisations engaged with supporting research.

O R C I D

Eloísa Rubio‐Beltrán https://orcid.org/0000-0002-2912-3632

Kristian A. Haanes https://orcid.org/0000-0001-5182-8957

R E F E R E N C E S

Abourashed, E. A., Koetter, U., & Brattström, A. (2004). In vitro binding experiments with a Valerian, Hops and their fixed combination extract (Ze91019) to selected central nervous system receptors.

Phytomedicine, 11, 633–638. https://doi.org/10.1016/j. phymed.2004.03.005

Alexander, S. P. H., Christopoulos, A., Davenport, A. P., Kelly, E., Marrion, N. V., Peters, J. A.,_{… CGTP Collaborators (2017). The Concise Guide to} PHARMACOLOGY 2017/18: G protein_{‐coupled receptors. British}

Jour-nal of Pharmacology, 174, S17_{–S129. https://doi.org/10.1111/} bph.13878

Barac, Y. D., Bar‐Am, O., Liani, E., Amit, T., Frolov, L., Ovcharenko, E., … Binah, O. (2012). I1 imidazoline receptor: Novel potential cytoprotective target of TVP1022, the S_{‐enantiomer of rasagiline. PLoS}

ONE, 7, e47890. https://doi.org/10.1371/journal.pone.0047890

Benemei, S., Cortese, F., Labastida_{‐Ramírez, A., Marchese, F., Pellesi, L.,} Romoli, M.,… School of Advanced Studies of the European Headache Federation (EHF_{‐SAS) (2017). Triptans and CGRP blockade—Impact} on the cranial vasculature. The Journal of Headache and Pain, 18, 103. https://doi.org/10.1186/s10194_{‐017‐0811‐5}

Bouchelet, I., Case, B., Olivier, A., & Hamel, E. (2000). No contractile effect for 5‐HT1Dand 5‐HT1Freceptor agonists in human and bovine cerebral arteries: Similarity with human coronary artery. British Journal of

Phar-macology, 129, 501_{–508. https://doi.org/10.1038/sj.bjp.0703081}

Bouchelet, I., Cohen, Z., Case, B., Séguéla, P., & Hamel, E. (1996). Differen-tial expression of sumatriptan_{‐sensitive 5‐hydroxytryptamine receptors} in human trigeminal ganglia and cerebral blood vessels. Molecular

Phar-macology, 50, 219–223.

van den Broek, R. W., MaassenVanDenBrink, A., Mulder, P. G., Bogers, A. J., Avezaat, C. J., John, G. W., & Saxena, P. R. (2002). Comparison of

contractile responses to donitriptan and sumatriptan in the human mid-dle meningeal and coronary arteries. European Journal of Pharmacology,

443, 125_{–132. https://doi.org/10.1016/S0014‐2999(02)01576‐5}

Buse, D. C., Reed, M. L., Fanning, K. M., Kurth, T., & Lipton, R. B. (2017). Cardiovascular events, conditions, and procedures among people with episodic migraine in the US population: Results from the American Migraine Prevalence and Prevention (AMPP) Study. Headache, 57, 31_{–44. https://doi.org/10.1111/head.12962}

Cason, B. A., Verrier, E. D., London, M. J., Mangano, D. T., & Hickey, R. F. (1987). Effects of isoflurane and halothane on coronary vascular resis-tance and collateral myocardial blood flowtheir capacity to induce coronary steal. Anesthesiology, 67, 665_{–675. https://doi.org/10.1097/} 00000542_{‐198711000‐00009}

Chan, K. Y., Baun, M., de Vries, R., van den Bogaerdt, A. J., Dirven, C. M., Danser, A. H.,… Gupta, S. (2011). Pharmacological characterization of VIP and PACAP receptors in the human meningeal and coronary artery.

Cephalalgia, 31, 181_{–189. https://doi.org/10.1177/0333102410} 375624

Chan, K. Y., de Vries, R., Leijten, F. P. J., Pfannkuche, H.‐J., van den Bogaerdt, A. J., Danser, A. H. J., & MaassenVanDenBrink, A. (2009). Functional characterization of contractions to tegaserod in human iso-lated proximal and distal coronary arteries. European Journal of

Pharmacology, 619, 61–67. https://doi.org/10.1016/j.ejphar.2009. 07.010

Chan, K. Y., Labruijere, S., Ramirez Rosas, M. B., de Vries, R., Garrelds, I. M., Danser, A. H.,_{… MaassenVanDenBrink, A. (2014). Cranioselectivity of} sumatriptan revisited: Pronounced contractions to sumatriptan in small human isolated coronary artery. CNS Drugs, 28, 273–278. https://doi. org/10.1007/s40263_{‐013‐0136‐0}

Chan, K. Y., Vermeersch, S., de Hoon, J., Villalón, C. M., & MaassenVanDenBrink, A. (2011). Potential mechanisms of prospective antimigraine drugs: A focus on vascular (side) effects. Pharmacology &

Therapeutics, 129, 332–351. https://doi.org/10.1016/j.pharmthera.

2010.12.001

Cohen, M. L., & Schenck, K. (2000). Contractile responses to sumatriptan and ergotamine in the rabbit saphenous vein: Effect of selective 5‐ HT1Freceptor agonists and PGF(2α). British Journal of Pharmacology,

131, 562_{–568. https://doi.org/10.1038/sj.bjp.0703587}

Colquhoun, D. (1998). Binding, gating, affinity and efficacy: The interpreta-tion of structure_{‐activity relationships for agonists and of the effects of} mutating receptors. British Journal of Pharmacology, 125, 924–947. https://doi.org/10.1038/sj.bjp.0702164

Curtis, M. J., Alexander, S., Cirino, G., Docherty, J. R., George, C. H., Giembycz, M. A.,_{… Ahluwalia, A. (2018). Experimental design and} anal-ysis and their reporting II: updated and simplified guidance for authors and peer reviewers. British Journal of Pharmacology, 175, 987–993. https://doi.org/10.1111/bph.14153

Cushing, D. J., Cooper, W. D., Gralinski, M. R., Lipicky, R. J., Kudenchuk, P. J., & Kowey, P. R. (2009). Comparison of the cardiac electrophysiology and general toxicology of two formulations of intravenous amiodarone in dogs. Cardiovascular Toxicology, 9, 126_{–133. https://doi.org/} 10.1007/s12012_{‐009‐9044‐4}

Diener, H._{‐C., & Limmroth, V. (2001). Advances in pharmacological} treat-ment of migraine. Expert Opinion on Investigational Drugs, 10, 1831–1845. https://doi.org/10.1517/13543784.10.10.1831 Dodick, D., Lipton, R. B., Martin, V., Papademetriou, V., Rosamond, W.,

MaassenVanDenBrink, A.,_{… The Triptan Cardiovascular Safety Expert} Panel (2004). Consensus statement: Cardiovascular safety profile of triptans (5‐HT agonists) in the acute treatment of migraine. Headache,

44, 414_{–425. https://doi.org/10.1111/j.1526‐4610.2004.04078.x}

Farkkila, M., Diener, H. C., Geraud, G., Lainez, M., Schoenen, J., Harner, N., … COL MIG‐202 study group (2012). Efficacy and tolerability of lasmiditan, an oral 5_‐HT1Freceptor agonist, for the acute treatment of migraine: A phase 2 randomised, placebo_{‐controlled, parallel‐group,} dose‐ranging study. Lancet Neurology, 11, 405–413.

Ferrari, M. D., Farkkila, M., Reuter, U., Pilgrim, A., Davis, C., Krauss, M.,… European COL_{‐144 Investigators (2010). Acute treatment of migraine} with the selective 5_‐HT1Freceptor agonist lasmiditan—A randomised proof‐of‐concept trial. Cephalalgia, 30, 1170–1178. https://doi.org/ 10.1177/0333102410375512

George, S. E., Bungay, P. J., & Naylor, L. H. (1997). Functional coupling of endogenous serotonin (5_‐HT1B) and calcitonin (C1a) receptors in CHO cells to a cyclic AMP‐responsive luciferase reporter gene. Journal

of Neurochemistry, 69, 1278–1285. https://doi.org/10.1046/j.1471‐

4159.1997.69031278.x

Goadsby, P. J., Wietecha, L. A., Dennehy, E. B., Kuca, B., Case, M. G., Aurora, S. K., & Gaul, C. (2019). Phase 3 randomized, placebo‐ controlled, double_{‐blind study of lasmiditan for acute treatment of} migraine. Brain, 142, 1894_{–1904. https://doi.org/10.1093/brain/} awz134

Harding, S. D., Sharman, J. L., Faccenda, E., Southan, C., Pawson, A. J., Ire-land, S., … NC‐IUPHAR (2018). The IUPHAR/BPS Guide to PHARMACOLOGY in 2018: Updates and expansion to encompass the new guide to IMMUNOPHARMACOLOGY. Nucleic Acids Research, 46, D1091–D1106. https://doi.org/10.1093/nar/gkx1121

Headache Classification Committee of the International Headache Society. (2018). The International Classification of Headache Disorders, 3rd edi-tion. Cephalalgia, 38, 1_–211.

Johnson, K. W., Schaus, J. M., Durkin, M. M., Audia, J. E., Kaldor, S. W., Flaugh, M. E.,_{… Phebus, L. A. (1997). 5‐HT}1Freceptor agonists inhibit neurogenic dural inflammation in guinea pigs. Neuroreport, 8,

2237_{–2239. https://doi.org/10.1097/00001756‐199707070‐00029} Kilkenny, C., Browne, W., Cuthill, I. C., Emerson, M., & Altman, D. G.

(2010). Animal research: Reporting in vivo experiments: The ARRIVE guidelines. British Journal of Pharmacology, 160, 1577_–1579. Kovalchin, J., Ghiglieri, A., Zanelli, E., Ings, R., & Mathers, T. (2016).

Lasmiditan acts specifically on the 5‐HT1Freceptor in the central ner-vous system. Cephalalgia, 36, 103.

Kuca, B., Silberstein, S. D., Wietecha, L., Berg, P. H., Dozier, G., Lipton, R. B., & COL MIG‐301 Study Group (2018). Lasmiditan is an effective acute treatment for migraine: A phase 3 randomized study. Neurology, 91 (24), e2222_–e2232.

Kurth, T., Winter, A. C., Eliassen, A. H., Dushkes, R., Mukamal, K. J., Rimm, E. B.,… Rexrode, K. M. (2016). Migraine and risk of cardiovascular dis-ease in women: Prospective cohort study. BMJ, 353, i2610.

Labruijere, S., Chan, K. Y., de Vries, R., van den Bogaerdt, A. J., Dirven, C. M., Danser, A. J.,_{… MaassenVanDenBrink, A. (2015).} Dihydroergota-mine and sumatriptan in isolated human coronary artery, middle meningeal artery and saphenous vein. Cephalalgia, 35, 182–189. https://doi.org/10.1177/0333102414544977

MaassenVanDenBrink, A., Reekers, M., Bax, W. A., Ferrari, M. D., & Saxena, P. R. (1998). Coronary side‐effect potential of current and pro-spective antimigraine drugs. Circulation, 98, 25_{–30. https://doi.org/} 10.1161/01.CIR.98.1.25

MaassenVanDenBrink, A., Reekers, M., Bax, W. A., & Saxena, P. R. (2000). Human isolated coronary artery contraction to sumatriptan characterised by the selective 5_‐HT1B/1D receptor antagonist GR55562. Pharmacology & Toxicology, 86, 287_{–290. https://doi.org/} 10.1111/j.0901_{‐9928.2000.860608.x}

MaassenVanDenBrink, A., van den Broek, R. W., de Vries, R., Bogers, A. J., Avezaat, C. J., & Saxena, P. R. (2000). Craniovascular selectivity of